React-promise-cache
enables you to cache promises and their associated data. This caching functionality operates uniformly on both the client and server sides. The library utilizes React-18 capabilities and is aimed towards the forthcoming advancements in React.
// Users list
import {useApi} from "react-promise-cache";
async function getUsersList() {
let promise = await API.get<UserType[]>(`/users`);
return promise.data
}
function UsersList() {
const usersList = useApi(getUsersList)
const data = React.use(usersList())
// or const data = usersList.use(); (falls back to React.use if existent)
// do something with data
}
And using parameters:
// User details
async function getUserDetailsList(id: number) {
let promise = await API.get<UserType>(`/users/${id}`);
return promise.data
}
function UserDetails({userId}: {userId: number}) {
const userDetails = useApi(getUserDetailsList)
const data = React.use(userDetails(userId))
// or const data = userDetails.use(); (falls back to React.use if existent)
// do something with data
}
This library uses react-18
features:
- When the
promise
ispending
, the component suspends and the nearest Suspense fallback will be shown. - When the
promise
is rejected, it is thrown to the nearestErrorBoundary
.
In the previous example, useApi
will return the same reference for the same
function reference. Ensure to memoize your function, but honestly, you won't
need since you can use parameters to your function.
Here is the different parts of the API that the library supports:
useApi
is designed to give you an Api
for your function, this Api has the
following properties:
Property | type | Description |
---|---|---|
(...args) |
(...args: A extends unknwon[]): State<T, R, A> or Promise<T> |
When it is invoked, it calls your original function and caches the result for further calls |
use(...args) |
use(...args: A extends unknwon[]): T |
Uses React.use or shims it to give you the resolved data. Will suspend on pending and throw on error. |
evict(...args) |
evict(...args: A extends unknwon[]): Api |
Removes the cache related to the given arguments |
subscribe(fn) |
subscribe(rerender: (() => void)): (() => void) |
Will subscribe to state changes and evicts |
getState(...args) |
getState(...args: A extends unknwon[]): State<T, R, A> |
Returns the cached state related to the given arguments |
useState(...args) |
useState(...args: A extends unknwon[]): State<T, R, A> |
Returns the cached state and performs subscription to state updates |
inject(fn) |
inject(fn: ((...args) => T or Promise<T>)): Api |
To lazily define your functions (supported with the app abstraction), you can inject the actual functions later |
So you can use this API with several React APIs. Either synchronous or async.
Here is a good example to try out!
The provider
is a React Context Provider that you should use to share, cache
and isolate your data.
It is optional in client side, but required in the server.
Property | type | Description |
---|---|---|
children |
React.ReactNode |
Your app |
cache? |
Map |
The actual cache map to use, it is created if not provided |
app? |
evict(...args: A extends unknwon[]): Api |
The return of createApp |
The Hydration
component is used to save your promises' data from the server
to the client, and boot from it there. It automatically caches the cache
instance in its context, and require a unique
string ID.
If you are streaming HTML and have multiple
Suspense
boundaries in your App, think of adding a Hydration every time you will suspend. Or useSuspendeBoundary
from the library (requiresid
as well) that injects aSuspense
and aHydration
, (you can pass thefallback
).
By @incepter, with 💜